Germany's manufacturing sector is investing billions in Industry 4.0 — and frequently failing at the same bottleneck: Operational Technology (OT) on the shop floor and Information Technology (IT) in the enterprise do not talk to each other. Proprietary protocols, decades-old PLC systems and strict availability requirements make integration the hardest problem in manufacturing digitalization. This article explains how AWS Outposts, IoT Core, IoT SiteWise and edge computing services build a viable OT/IT architecture for manufacturing — without production stoppages and without risky big-bang migrations. Target audience: IT directors, OT managers and digitalization leads at manufacturing companies.
The OT/IT Gap: Why Industry 4.0 Projects Stall
Industry 4.0 promises connected factories, real-time transparency and data-driven decisions. The reality in German manufacturing plants often looks different: PLCs (Programmable Logic Controllers) communicate over proprietary protocols like Profibus or Modbus. SCADA systems run on isolated networks with no internet access. Data from the production floor ends up in spreadsheets at best, and nowhere at all at worst.
A 2024 VDMA study found that while 78% of German machinery manufacturers are engaged with digitalization, only 23% have achieved end-to-end data integration from sensor to ERP system. The gap between aspiration and reality is wide.
The core problem: OT and IT were built over decades with different priorities. OT prioritizes availability and determinism — a PLC must not crash; response times must be predictable. IT prioritizes flexibility, scalability and connectivity. AWS bridges these two worlds with a hybrid approach.
Key Terms: OT, IT, Edge and Cloud
- Operational Technology (OT)
- Hardware and software that monitors and controls physical devices, processes and events. This includes PLCs, SCADA systems, DCS (Distributed Control Systems) and industrial sensors. OT is designed for real-time capability and high availability.
- Information Technology (IT)
- Systems for processing, storing and transmitting information: ERP, CRM, BI platforms, cloud services. IT is optimized for scalability, flexibility and connectivity.
- OT/IT Convergence
- The integration of both worlds into a continuous data flow — from sensor to decision. The goal is to make OT data available in IT systems without compromising the stability and security of the OT layer.
- Edge Computing
- Processing data as close as possible to where it is generated — in the factory hall rather than in the cloud. Edge computing reduces latency, saves bandwidth and enables local decisions even during WAN outages.
- OPC UA (OPC Unified Architecture)
- Industry standard for machine-readable data description and communication between industrial assets. OPC UA is platform-independent, secure and the preferred protocol for OT/IT integration on AWS.
The AWS Architecture for Industry 4.0
AWS provides a layered model that accommodates the realities of manufacturing environments. The architecture is organized into three tiers:
| Layer | Location | AWS Service | Function |
|---|---|---|---|
| Edge Layer | Shop floor / PLC network | AWS IoT Greengrass | Local processing, ML inference, MQTT broker, OPC UA connectivity |
| On-Premises Cloud | Factory data center / server room | AWS Outposts | AWS APIs on-premises, low-latency processing, local data residency |
| Cloud Layer | AWS Region (Frankfurt eu-central-1) | IoT Core, IoT SiteWise, Kinesis, S3, SageMaker | Scalable data ingestion, historical analysis, ML training, reporting |
AWS IoT Greengrass: Edge Intelligence
AWS IoT Greengrass is the entry point for most manufacturing companies. Greengrass is installed as a software agent on a local gateway machine and securely connects OT devices to the cloud. Key capabilities: MQTT broker for protocol translation, local Lambda functions for data filtering and normalization, ML inference at the edge — even without a WAN connection — and a native OPC UA connector for production line integration.
AWS Outposts: Cloud Infrastructure on the Shop Floor
For scenarios with hard latency requirements (under 5 ms) or regulatory requirements for local data residency, AWS Outposts is the right choice. Outposts brings native AWS infrastructure into the factory data center: the same APIs, the same services, the same consistency as in the AWS Region — but operated locally.
AWS IoT SiteWise: Structuring and Analyzing Industrial Data
AWS IoT SiteWise is purpose-built for industrial time-series data. SiteWise models the factory as an asset hierarchy (Factory → Production Line → Machine → Sensor), collects OPC UA data via the SiteWise Edge Gateway and makes it available for analysis and dashboards. Particularly valuable: SiteWise automatically calculates OEE (Overall Equipment Effectiveness) based on raw data.
Implementation Roadmap: OT/IT Integration in Four Phases
- Assessment & Pilot Line (4–8 weeks): Inventory of all OT systems (PLCs, SCADA, sensors, protocols). Selection of a pilot line with clear business value. Installation of AWS IoT Greengrass on a gateway machine. First OPC UA connection to a machine. Data visible in SiteWise.
- Pilot Line in Production (4–6 weeks): Build SiteWise asset model for the pilot line. Create OEE dashboard. Train first anomaly detection model with Amazon Lookout for Equipment. Stabilize operations and collect feedback from production staff.
- Rollout to Additional Lines (3–6 months): Apply proven architecture pattern to additional production lines and sites. Build central asset model of the entire factory. Migrate historical data from legacy systems into the data lake.
- Advanced Analytics & AI (ongoing): Train and deploy SageMaker models for predictive maintenance, quality prediction and production optimization. Build digital twin. Integrate with ERP/MES for closed-loop control.
OEE as the Key Metric
Overall Equipment Effectiveness (OEE) is the central efficiency metric in manufacturing. OEE consists of three factors: Availability, Performance and Quality. AWS IoT SiteWise calculates OEE automatically based on incoming sensor data.
| OEE Component | Definition | SiteWise Calculation |
|---|---|---|
| Availability | Share of planned production time the asset is actually running | Run time / (run time + downtime), from PLC status signal |
| Performance | Ratio of actual to theoretical maximum production rate | Actual cycle / ideal cycle, from cycle counter signal |
| Quality | Share of good parts in total production | Good parts / total parts, from quality sensor or MES feedback |
| OEE | Availability × Performance × Quality | Automatically calculated, historically queryable, per line and asset |
World-class OEE starts at 85%. The average in German manufacturing sits at 55–65% according to VDMA. A 10-percentage-point improvement often translates to millions of euros in additional output per year for a mid-sized production line.
Storm Reply: Industry 4.0 for German Manufacturing
Storm Reply is an AWS Premier Consulting Partner with strong experience in manufacturing and industrial projects. As part of the Reply Group, Storm Reply combines the industrial know-how of Cluster Reply (OT expertise) with deep AWS competency across IoT, Migration, Machine Learning and Cloud Operations.
Storm Reply guides manufacturing companies from assessment through pilot to full-scale implementation — with a clear focus on rapid business value and minimal production disruption.
Frequently Asked Questions
- What is OT/IT convergence in manufacturing?
- OT/IT convergence is the integration of Operational Technology (production assets, PLCs, SCADA) with Information Technology (ERP, cloud, analytics). The goal is end-to-end data flow from sensor to decision — without media breaks and proprietary silos.
- Which AWS services are suited for Industry 4.0?
- AWS IoT Greengrass for edge computing and protocol translation, AWS IoT Core for scalable device connectivity, AWS IoT SiteWise for industrial time-series analysis and OEE calculation, AWS Outposts for low-latency on-premises processing, Amazon SageMaker for ML models.
- How long does an OT/IT integration with AWS take?
- A proof of concept with one production line is achievable in 6–8 weeks. Full factory integration takes 6–18 months depending on the variety of assets and legacy systems. AWS MAP funds up to 50% of migration costs.
- Is AWS Outposts suitable for every manufacturing environment?
- AWS Outposts is optimal for scenarios with latency requirements below 5 ms or regulatory requirements for local data residency. For less time-critical data, AWS IoT Greengrass often suffices as a lightweight edge solution.
OT/IT Integration Assessment
Let us evaluate together which AWS architecture fits your manufacturing environment — from pilot line to smart factory.
Get in touch